Advertisement

On the UTA Methods for Solving the Model Selection Problem

  • Valentina MinnettiEmail author
Conference paper
Part of the Springer Proceedings in Mathematics & Statistics book series (PROMS, volume 217)

Abstract

In this paper, the multiple criteria UTA methods are proposed for solving the problem of model selection. The UTA methods realize a ranking of the models, from the best one to the worst one, by means of the comparisons of their global utility values. These values are computed by means of Linear Programming problems. Two UTA methods are illustrated. An example, that examines the performances of some classification models in the web context, is presented.

Keywords

UTA methods Linear programming Classification models 

References

  1. 1.
    Kadane, J.B. Lazar, N.A.: Methods and criteria for model selection. J. Am. Stat. Assoc. 99(465) (2004)Google Scholar
  2. 2.
    Zucchini, W.: An introduction to model selection. J. Math. Psychol. 44, 41–61 (2000)CrossRefzbMATHGoogle Scholar
  3. 3.
    Mousseau, V., Slowinski, R., Zielniewicz, P.: ELECTRE TRI 2.0, a methodological guide and users manual. Document du LAMSADE no111, Universit Paris-Dauphine (1969)Google Scholar
  4. 4.
    Siskos, Y., Grigoroudis, E., Matsatsinis, N.: Multiple criteria decision analysis: state of the art survey. In: Figueira, J., Greco, S., Ehrgott, M.(eds.), pp. 297–344. Springer, Boston, Dordrecht, London (2005)Google Scholar
  5. 5.
    Barcaroli, G., Bianchi, G., Bruni, R. Nurra, A., Salomone, S., Scarn\(\acute{o}\), M.: Machine learning and statistical inference: the case of Istat survey on ICT. In: Proceedings of the 48th Scientific Meeting of the Italian Statiswtical Society, Salerno, June 2016Google Scholar
  6. 6.
    Bruni, R., Bianchi, G.: Effective Classification using Binarization and Statistical Analysis. IEEE Trans. Knowl. Data Eng. 9, 2349–2361 (2015)CrossRefGoogle Scholar
  7. 7.
    Guitouni, A., Martel, J.M.: A multidimensional evaluation of the effectiveness of business incubator: an application of the PROMETHEE outranking method. Eur. J. Oper. Res. 109, 501–521 (1998)CrossRefGoogle Scholar
  8. 8.
    Greco, S., Mousseau, V., Slowinski, R.: Ordinal regression revisited: multiple criteria ranking using a set of additive value functions. Eur. J. Oper. Res. 191, 415–435 (2008)CrossRefzbMATHMathSciNetGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Faculty of Information Engineering, Department of Statistic Science, Informatics and StatisticsSapienza University of RomeRomeItaly

Personalised recommendations